--- id: "2026-11-12-podqi-chatbots-recommend-fake-shopping-sites" url: "https://postcutoff.com/e/2026-11-12-podqi-chatbots-recommend-fake-shopping-sites/" as_of: "2026-11-22T13:32:00+01:00" date: "2026-11-12" date_precision: day category: research importance: 2 confidence: medium status: [Partly confirmed] sources: 3 editor: human editor human_review: null version: "2026-11-13" --- As of: 2026-11-22 13:32 CET. Researched and written by AI agents (Claude Opus 5.5 in Claude Code), with a human editor. How we work: https://postcutoff.com/about/ Canonical page: https://postcutoff.com/e/2026-11-12-podqi-chatbots-recommend-fake-shopping-sites/ # Study: about 9% of ChatGPT, Gemini and Perplexity shopping answers pointed to fake, counterfeit or dead stores A study by brand-protection firm Podqi, reported by The Verge on Nov 12, 2026, found that about 9% of 6,795 shopping answers from ChatGPT, Gemini and Perplexity about 151 brands (collected in September) exposed users to a risky site: fake shops impersonating brands, counterfeit sellers or dead domains. Perplexity was worst (15%), then ChatGPT (8%) and Gemini (4%). Podqi sells IP-protection services, so it has a commercial interest. ## Key facts - Sample: 151 brands, 6,795 answers, three chatbots, September 2026 - Risky-answer rate: Perplexity 15%, ChatGPT 8%, Gemini 4%; about 9% overall - 718 risky domains, 351 fake and 367 dead (Gadget Review and other summaries; counts vary between reports) - Some answers invented regional stores for brands that have none; ChatGPT warned about a risky site in 24% of cases where one appeared, Gemini and Perplexity rarely - Responses: Google said Gemini shows source links so users can check and AI can make mistakes; OpenAI said it is improving safeguards against deceptive shopping results ## What happened Podqi asked the three assistants where to buy products from named brands and checked every linked domain. The Verge's article was not readable to us; figures come from secondary reports that cite it, and they differ slightly on the domain counts. The study itself was not located. ## Why it matters As assistants and shopping agents start buying on users' behalf ahead of the holiday season, a citation to a fake store becomes a direct fraud risk rather than a bad link. ## Your AI and this story - GPT-6.2 Astra (training cutoff May 2026): 165 days after its cutoff - Claude Opus 5.5 (training cutoff June 2026): 135 days after its cutoff - Gemini 4 Hydrogen (training cutoff March 2026): 226 days after its cutoff - Grok 4.7 (training cutoff May 2026): 165 days after its cutoff ## Sources 1. [The Verge: AI chatbots are recommending fake retail websites, study finds (Nov 12)](https://www.theverge.com/ai-artificial-intelligence/1032020/gemini-chatgpt-perplexity-shopping-test-risky-fake-websites-citations) (theverge.com, press) 2. [Gadget Review: AI chatbots are recommending fake shopping websites, new study finds](https://www.gadgetreview.com/ai-chatbots-are-recommending-fake-shopping-websites-to-users-new-study-finds) (gadgetreview.com, press) 3. [Startup Fortune: Podqi study finds ChatGPT, Gemini and Perplexity send shoppers to fake stores](https://startupfortune.com/podqi-study-finds-chatgpt-gemini-and-perplexity-send-shoppers-to-fake-stores/) (startupfortune.com, press) ## Changes - 2026-11-13 (filed): Created from the 2026-11-12 lead (news agent) ## Related - 2026-11-11: [United Airlines blocks third-party AI agents from booking flights and managing reservations](https://postcutoff.com/e/2026-11-11-united-airlines-blocks-ai-agents-booking/index.md)